Low-Resource Cross-Lingual Summarization through Few-Shot Learning with Large Language Models

TL;DR

This paper explores the effectiveness of few-shot learning with large language models for low-resource cross-lingual summarization, demonstrating significant improvements with GPT-3.5 and GPT-4 but limited success with open-source models.

Contribution

It is the first comprehensive study on few-shot XLS performance of various LLMs, highlighting the potential and challenges of this approach for low-resource languages.

Findings

GPT-3.5 and GPT-4 significantly improve XLS performance with few-shot learning.

Open-source Mistral-7B-Instruct-v0.2 struggles with limited examples.

Few-shot learning shows promise but requires further research for optimal results.

Abstract

Cross-lingual summarization (XLS) aims to generate a summary in a target language different from the source language document. While large language models (LLMs) have shown promising zero-shot XLS performance, their few-shot capabilities on this task remain unexplored, especially for low-resource languages with limited parallel data. In this paper, we investigate the few-shot XLS performance of various models, including Mistral-7B-Instruct-v0.2, GPT-3.5, and GPT-4. Our experiments demonstrate that few-shot learning significantly improves the XLS performance of LLMs, particularly GPT-3.5 and GPT-4, in low-resource settings. However, the open-source model Mistral-7B-Instruct-v0.2 struggles to adapt effectively to the XLS task with limited examples. Our findings highlight the potential of few-shot learning for improving XLS performance and the need for further research in designing LLM architectures and pre-training objectives tailored for this task. We provide a future work direction to explore more effective few-shot learning strategies and to investigate the transfer learning capabilities of LLMs for cross-lingual summarization.